zlacker

[return to "MobileDiffusion: Rapid text-to-image generation on-device"]
1. minima+s9[view] [source] 2024-01-31 23:36:39
>>jasond+(OP)
> With superior efficiency in terms of latency and size, MobileDiffusion has the potential to be a very friendly option for mobile deployments given its capability to enable a rapid image generation experience while typing text prompts. And we will ensure any application of this technology will be in-line with Google’s responsible AI practices.

So I'm interpreting this that it won't ever get released.

◧◩
2. Legend+de[view] [source] 2024-02-01 00:13:22
>>minima+s9
I would interpret it as "expect to see this powering some features in the next-generation Pixel".
◧◩◪
3. jshear+df[view] [source] 2024-02-01 00:21:23
>>Legend+de
We've already seen this progression - they debuted Magic Eraser as a cloud feature, then with the Pixel 8 they got it running locally on the device. But they also introduced Magic Editor with the Pixel 8, running on the cloud, and the Pixel 9 or 10 will probably run it on-device.
◧◩◪◨
4. ollin+EL[view] [source] 2024-02-01 06:59:38
>>jshear+df
it may turn out more like the imagen timeline

2022-05 - google imagen research paper posted >>31484562

2022-12 - imagen developers leave google to form ideogram

2023-08 - ideogram ships a version of imagen, free, for anyone who wants to use it https://ideogram.ai/publicly-available

2023-12 - google "imagen 2" is officially "generally available for Vertex AI customers on the allowlist (i.e., approved for access)." >>38628417

[go to top]